Pose / position detection device for a detector

Abstract

A posture / position detection device (10) provided with a detector (50) mounted on a vehicle (100, 100a) is provided. The posture / position detection device (10) includes an image acquisition unit (21) that acquires a distance image (KI) and a background light image (HI) from the detector, the distance image representing a distance to an object within an irradiation range of irradiation light emitted from the detector, the background light image being a background light image representing a light receiving intensity of ambient light and having a resolution higher than that of the distance image, and a posture / position detection unit (23) that detects a posture / position of the detector using the distance image and the background light image.

Description

[0001] Cross-references to related applications

[0002] This application claims priority based on Japanese Patent Application No. 2020-032762, filed on February 28, 2020, the entire disclosure of which is incorporated herein by reference. Technical Field

[0003] This disclosure relates to techniques for detecting the posture and position of a detector mounted on a vehicle. Background Technology

[0004] In order to obtain appropriate detection results from detectors mounted on vehicles, a technique has been proposed that fuses the detection results of two detectors of different types and resolutions to determine calibration parameters such as the orientation and position of each detector (e.g., International Publication No. 2017-122529).

[0005] However, in the technology described in International Publication No. 2017-122529, due to the different types of the two detectors, there are cases where the detection ranges of the detectors do not overlap, or even if the detection ranges overlap, the overlap is small. In such cases, it is possible that the detection results available for determining the calibration parameters are reduced, making it impossible to determine the calibration parameters with good accuracy. Therefore, a technology is desired that can accurately detect at least one of the orientation and position of the detector mounted on the vehicle. Summary of the Invention

[0006] This disclosure can be implemented in the following ways.

[0007] According to one embodiment of this disclosure, a posture / position detection device for a detector mounted on a vehicle is provided. The posture / position detection device includes: an image acquisition unit that acquires a distance image and a background light image from the detector at different time intervals, the distance image representing the distance to an object within the illumination range of illumination light emitted from the detector, and the background light image being a background light image representing the intensity of ambient light, and having a resolution higher than that of the distance image; and a posture / position detection unit that uses the distance image and the background light image to detect the posture / position of the detector.

[0008] According to this method, the detector pose / position detection device acquires a distance image and a background light image with a resolution higher than that of the distance image from the detector, and uses the distance image and the background light image to detect the detector pose / position. Therefore, it is possible to detect the detector pose / position using only the detector's detection data. Thus, compared to structures that use detection data from two different types of detectors to detect the detector pose / position, it is possible to detect the detector pose / position with high accuracy.

[0009] The present disclosure can also be realized in various ways. For example, it can also be realized as a posture / position detection method of a detector, a posture / position detection program of a detector, or a recording medium readable by a computer on which the program is recorded. BRIEF DESCRIPTION OF DRAWINGS

[0010] The above and other objects, features and advantages of the present disclosure will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0011] Figure 1 is a diagram illustrating a schematic configuration of a posture / position detection system of a detector that is an embodiment of the present disclosure,

[0012] Figure 2 is a diagram illustrating a schematic configuration of a posture / position detection system of a detector,

[0013] Figure 3 is a diagram illustrating one example of a target,

[0014] Figure 4 is a block diagram schematically illustrating a schematic configuration of a posture / position detection device,

[0015] Figure 5 is a diagram illustrating a configuration of a light-receiving element,

[0016] Figure 6 is a diagram schematically illustrating one example of a background light image,

[0017] Figure 7 is a diagram schematically illustrating one example of a distance image,

[0018] Figure 8 is a flowchart illustrating a process procedure of a posture / position detection process,

[0019] Figure 9 is a diagram illustrating a corresponding pixel and surrounding pixels,

[0020] Figure 10 is a flowchart illustrating a process procedure of a posture / position detection process of the second embodiment,

[0021] Figure 11 is a diagram illustrating one example of a target of the second embodiment,

[0022] Figure 12 is a diagram illustrating a schematic configuration of a posture / position detection system of a detector of the third embodiment,

[0023] Figure 13is a flowchart showing a process procedure of the posture / position detection process of the third embodiment,

[0024] Figure 14 is an explanatory view showing another example of the target,

[0025] Figure 15 is an explanatory view showing another example of the target,

[0026] Figure 16 is an explanatory view showing another example of the target,

[0027] Figure 17 is an explanatory view showing another example of the target. DETAILED DESCRIPTION

[0028] First Embodiment:

[0029] As shown in Figure 1 and Figure 2 , the posture / position detection system 500 of the detector 50 mounted on the vehicle 100 is provided with a plurality of targets TG, the posture / position detection device 10, and the measuring apparatus 300. The plurality of targets TG are physical targets used for detection of at least either of the posture and the position of the detector 50, and are arranged around the vehicle 100. In the posture / position detection system 500, the coordinate positions of the target TG in the three-dimensional space determined by one or more feature points on the target TG are detected by the posture / position detection device 10 and the measuring apparatus 300, respectively, and the offset amount of the two coordinate positions is calculated, whereby the posture / position of the detector 50 with respect to the reference coordinate axis described later is detected. In the present specification, the description of the posture / position means at least either of the posture and the position.

[0030] As shown in Figure 2 , the plurality of targets TG are planar plates on which a predetermined detection pattern KP is drawn on the surface. The surface of the target TG opposite to the vehicle 100 is provided with the detection pattern KP in a checkered pattern. As shown in Figure 3 , the detection pattern KP is composed of a plurality of patterns in a lattice shape, and a black pattern Pt1 and a white pattern Pt2 are alternately arranged. At the intersection of each pattern Pt1 and Pt2, a mark indicating a feature point Fp is displayed. Since the black pattern Pt1 and the white pattern Pt2 are adjacent to each other, in other words, the patterns Pt1 and Pt2 having different luminances are adjacent to each other, edge detection in the background light image described later becomes easy, and the pixel position of the feature point Fp on the background light image can be detected with high accuracy.

[0031] The posture / position detection device 10 detects the posture / position of the detector 50 by executing a posture / position detection process described later. Specifically, the posture / position detection device 10 calculates the distance of the feature point Fp on the target TG using the detection data acquired from the detector 50, specifically, the background light image and the distance image described later, and converts the calculated distance into a three-dimensional position. Then, the posture / position detection device 10 detects the posture / position of the detector 50 with respect to the reference coordinate axis described later by comparing the converted three-dimensional position with the three-dimensional position acquired by the measuring device 300.

[0032] The posture / position detection device 10 decides the coordinate axis serving as the reference of the detection of the posture / position of the detector 50 using the position of the vehicle 100 acquired by the measuring device 300 (hereinafter, referred to as "reference coordinate axis"). In addition, the posture / position detection device 10 decides the three-dimensional position of the feature point Fp of the target TG with respect to the reference coordinate axis (hereinafter, referred to as "reference position") using the position of the feature point Fp on the target TG acquired by the measuring device 300. Specifically, the posture / position detection device 10 detects the positions and directions of the four wheels of the vehicle 100 using the measuring device 300 to decide the direction of the reference coordinate axis. The posture / position detection device 10 detects the three-dimensional position (X, Y, Z) of the feature point Fp on the target TG with respect to the reference coordinate axis using the measuring device 300 to measure the feature point Fp of each target TG. The posture / position detection device 10 sets the three-dimensional position of the feature point Fp detected by the measuring device 300 as the reference position. In the present embodiment, the measuring device 300 is, for example, a total station. Further, based on the specification table of the vehicle 100 and the position of the target in front of the vehicle 100 acquired by the measuring device 300, the origin position of the reference coordinate axis is offset to the front end of the vehicle 100. The origin position of the reference coordinate axis is not limited to the front end of the vehicle 100, and can be set to, for example, the position of the rear wheel shaft center of the vehicle 100.

[0033] Detector 50 is mounted on the roof at the front of vehicle 100, illuminating the object with illumination light and receiving reflected light from the object. In addition to reflected light from the object, detector 50 also receives other types of light, such as sunlight, streetlights, headlights from other vehicles, and ambient light (hereinafter also referred to as "background light") reflected from the object. Detector 50 determines the light received after removing the background light as reflected light from the object and determines the time from illuminating the object to receiving the reflected light, i.e., the time of flight (TOF). Detector 50 calculates the TOF as the distance to the object. Therefore, the attitude / position detection device 10 acquires a reflection intensity image (using the intensity of the reflected light as pixel values ​​for each pixel), a background light image (using the intensity of the background light as pixel values ​​for each pixel), and a distance image (representing the distance to the object), and uses these three images to detect the object. In this embodiment, detector 50 is a LiDAR (Light Detection and Ranging) system. In addition, the detector 50 can also be configured on the front grille, front window, front bumper, rear window, rear bumper, front fender, and rear fender of the vehicle 100, and multiple detectors can also be provided.

[0034] Although Figure 1 as well as Figure 2 The diagram is omitted, but the vehicle 100 may also include a control device for performing driving assistance or automatic driving of the vehicle 100, a data processing device for processing detection data input from the detector 50, etc. In addition, the vehicle 100 may also include a posture / position detection device 10.

[0035] like Figure 4 As shown, the attitude / position detection device 10 is composed of a personal computer having a CPU 20, a memory 30, and an input / output interface 11. The CPU 20, memory 30, and input / output interface 11 are connected via a bus 15 in a manner capable of bidirectional communication. The memory 30 includes ROM, RAM, and EEPROM. The light emission control unit 51 and light receiving control unit 55 of the detector 50 are respectively connected to the input / output interface 11 via control signal lines.

[0036] The light emission control unit 51 drives the light emission element 52 to emit illumination light at a timing corresponding to the light emission command signal input from the posture / position detection device 10. The light emission element 52 is, for example, an infrared laser diode, which emits infrared laser light as illumination light. There can be one or more light emission elements 52. The illumination light emitted by the light emission element 52 is reflected by an object existing within the illumination range of the detector 50. The light, including the reflected light from the object and the background light, is received by the light receiving element 56.

[0037] The light-receiving control unit 55 acquires the light-receiving value of the incident light incident on the light-receiving element 56 for each pixel according to the light-receiving command signal input from the posture / position detection device 10, and outputs a background light image HI, a distance image KI, and a reflection intensity image RI. In this embodiment, "background light image HI" means an image that uses the brightness of the light received by the light-receiving element 56 as a pixel value when the light is not emitted from the light-emitting element 52. For example, the background light image HI can be acquired before the light-emitting control unit 51 drives the light-emitting element 52. "Distance image KI" means an image that uses the distance to a detection point (detection point group) that indicates the location of at least a part of the object determined based on the reflected light as a pixel value. "Reflection intensity image RI" means an image that uses the brightness of the light received by the light-receiving element 56 as a pixel value, representing the amount of light emitted by the light-emitting element 52 and reflected by the object.

[0038] like Figure 5 As shown, the light-receiving element 56 is an array of light-receiving elements 58 arranged in a planar manner on the light-receiving surface 57. The light-receiving elements 58 are, for example, composed of a single-photon avalanche diode (SPAD). Other types of light-receiving elements, such as PIN photodiodes or APDs, can also be used as light-receiving elements 58. The light-receiving control unit 55 treats a set of light-receiving elements 58 in the horizontal direction (H) and in the vertical direction (V) as a pixel, divides the multiple light-receiving elements 58 arranged on the light-receiving surface 57 into multiple pixels Ps arranged in a two-dimensional manner, and outputs the light-receiving results of the [H×V] light-receiving elements 58 contained in each pixel Ps as the pixel value of each pixel Ps. H and V are integers greater than or equal to 1. The number of light-receiving elements 58 constituting a pixel Ps [H×V] is also called the "pixel size". The smaller the pixel size, the higher the resolution (deg) of the pixel detecting light including reflected light containing the irradiated light, and the higher the resolution of the acquired image.

[0039] The pixel size of a pixel Ps, i.e., the number of light-receiving elements 58 [H×V], is preset by the light-receiving control unit 55. In this embodiment, the resolution of the background light image HI is set to be higher than the resolution of the distance image KI and the reflection intensity image RI. Specifically, Figure 5The pixel Ps_h, indicated by the dashed box, represents a pixel in the background light image HI with a pixel size set to H=V=2. The pixel Ps_k, indicated by the dotted-line box, represents a pixel in the distance image KI and the reflection intensity image RI, both with pixel sizes set to H=4 and V=6. Therefore, the resolution of pixel Ps_h is higher than that of pixel Ps_k. The output is an image whose pixel value is determined by the light-receiving values ​​of the [H×V] light-receiving elements 58 contained in each pixel Ps. Therefore, the output is a background light image HI with a higher resolution than the distance image KI.

[0040] Figure 6 The background light image HI for the target TG is shown. Figure 7 The distance image KI for the target TG is shown. Pixel G1 of the background light image HI is composed of four light-receiving elements 58 contained in pixel Ps_h. Pixel G2 of the distance image KI is composed of six light-receiving elements 58 contained in pixel Ps_k. Therefore, the number of pixels R in the horizontal direction in the background light image HI is greater than the number of pixels M in the horizontal direction in the distance image KI, and the number of pixels Q in the vertical direction in the background light image HI is greater than the number of pixels N in the vertical direction in the distance image KI. Furthermore, the pixel sizes of pixels Ps_k and Ps_h are just examples and are not limited to; any other arbitrary values ​​satisfying 0 < Ps_h < Ps_k can be set.

[0041] like Figure 4 As shown, the CPU 20 functions as the image acquisition unit 21, the pose / position detection unit 23, and the calibration unit 25 by unfolding and executing the program stored in the memory 30.

[0042] The image acquisition unit 21 acquires a background light image HI, a distance image KI, and a reflection intensity image RI from the detector 50. The pose / position detection unit 23 uses the background light image HI and the distance image KI to compare the three-dimensional position of the detected feature point Fp with a reference position, thereby detecting the pose / position of the detector 50 relative to the reference coordinate axes. The calibration unit 25 calibrates the pose / position of the detector 50 to compensate for any offset between the pose / position of the detector 50 detected by the pose / position detection unit 23 and the predetermined pose / position of the detector 50 used as a reference.

[0043] This is performed by CPU 20 executing a pose / position detection program (not shown) pre-stored in memory 30. Figure 8 The pose / position detection processing of the detector is shown. For example, on the inspection line of vehicle 100 before leaving the factory, pose / position detection processing is performed in order to adjust the pose / position of detector 50.

[0044] The posture / position detection section 23 acquires the reference position (step S105). Specifically, the posture / position detection section 23 acquires the reference position by detecting the feature points Fp on the target TG using the measuring device 300. Further, the posture / position detection section 23 can store the acquired reference position in the memory 30.

[0045] The image acquisition section 21 acquires the background light image HI from the detector 50 (step S110). The image acquisition section 21 acquires the distance image KI from the detector 50 (step S115).

[0046] The posture / position detection section 23 detects the pixel position of the feature point Fp on the target TG in the background light image HI (step S120). Specifically, the posture / position detection section 23 extracts the pixels of the corner portion of the target TG from the background light image HI, for example, using the corner detection method of Harris.

[0047] The posture / position detection section 23 detects the position of the pixel (hereinafter, referred to as "corresponding pixel") in the distance image KI corresponding to the pixel position of the feature point Fp on the target TG in the background light image HI, and extracts the pixels around the corresponding pixel (hereinafter, referred to as "surrounding pixels") (step S125). Specifically, the posture / position detection section 23 first extracts the corresponding pixel Gt (the pixel shown by the dotted line) of the feature point Fp1, for example, from the distance image KI. Next, the posture / position detection section 23 extracts the pixels Gs1, Gs2, Gs3, Gs4, Gs5, Gs6, Gs7, and Gs8 (the pixels shown by the dash-dotted line) surrounding the corresponding pixel Gt and adjacent to the corresponding pixel Gt, as the surrounding pixels Gs of the corresponding pixel Gt. Similarly, the corresponding pixel Gt and the surrounding pixels Gs are extracted also for the other feature points Fp (Fp2, Fp3, and Fp4 in the example shown in FIG. 6) on the target TG. Figure 9 Figure 9 The posture / position detection section 23 detects the position of the pixel (hereinafter, referred to as "corresponding pixel") in the distance image KI corresponding to the pixel position of the feature point Fp on the target TG in the background light image HI, and extracts the pixels around the corresponding pixel (hereinafter, referred to as "surrounding pixels") (step S125). Specifically, the posture / position detection section 23 first extracts the corresponding pixel Gt (the pixel shown by the dotted line) of the feature point Fp1, for example, from the distance image KI. Next, the posture / position detection section 23 extracts the pixels Gs1, Gs2, Gs3, Gs4, Gs5, Gs6, Gs7, and Gs8 (the pixels shown by the dash-dotted line) surrounding the corresponding pixel Gt and adjacent to the corresponding pixel Gt, as the surrounding pixels Gs of the corresponding pixel Gt. Similarly, the corresponding pixel Gt and the surrounding pixels Gs are extracted also for the other feature points Fp (Fp2, Fp3, and Fp4 in the example shown in FIG. 6) on the target TG. Figure 9 ​In this context, the surrounding pixel Gs is at most one pixel (a total of eight pixels) adjacent to the corresponding pixel Gt, but it can also be set to at most two pixels (a total of twenty-four pixels) or at most three pixels (a total of forty-eight pixels) adjacent to the corresponding pixel Gt.

[0048] like Figure 8 As shown, the pose / position detection unit 23 acquires the distance of each pixel of the surrounding pixel Gs (step S130). Specifically, the pose / position detection unit 23 acquires the pixel values ​​of each of the surrounding pixels Gs1, Gs2, Gs3, Gs4, Gs5, Gs6, Gs7, and Gs8 on the distance image KI. The acquired distance of each surrounding pixel Gs is equivalent to the "distance of surrounding points" in this disclosure.

[0049] The pose / position detection unit 23 converts the acquired distances to surrounding points into three-dimensional coordinates (step S135). Specifically, the pose / position detection unit 23 performs a known coordinate transformation process that converts polar coordinates to orthogonal coordinates. As a result, the positions of the corresponding pixel Gt and the surrounding pixels Gs on the distance image KI in three-dimensional space (hereinafter also referred to as "three-dimensional positions of surrounding points") can be obtained.

[0050] The pose / position detection unit 23 calculates the three-dimensional position of the feature point Fp using the three-dimensional positions of the surrounding pixels Gs (step S140). Specifically, since the target TG is planar, the pose / position detection unit 23 performs interpolation processing using the three-dimensional positions of pixels in the surrounding pixels Gs obtained from the target TG to calculate the three-dimensional position of the feature point Fp (hereinafter also referred to as "corresponding point three-dimensional position"). Whether a surrounding pixel Gs is a pixel obtained from the target TG is determined if the difference in distance between the corresponding pixel Gt and each surrounding pixel Gs in three-dimensional space is a predetermined threshold. The threshold is, for example, 0.3 meters. In this embodiment, the corresponding point three-dimensional position is calculated by interpolation processing, so compared to the structure that calculates the three-dimensional position of the feature point Fp using the three-dimensional position calculated from the corresponding pixel Gt on the distance image KI, the three-dimensional position of the feature point Fp can be calculated with high precision (sub-pixel precision). Furthermore, as the interpolation processing, for example, nearest neighbor method, bilinear method, and bicubic method can be used.

[0051] The posture / position detecting section 23 detects the posture / position of the detector 50 (step S145). Specifically, the posture / position detecting section 23 uses the respective three-dimensional positions, gradient information, and the like, and uses the nearest neighbor method, the Global Nearest Neighbor method to establish correspondence between the three-dimensional positions of the plurality of feature points Fp acquired by the detector 50 and the plurality of reference positions acquired by the measuring device 300, and calculates the posture and the position of the detector 50 with respect to the reference coordinate axes using four or more corresponding points by the following expression (1).

[0052] X' = RX + T... (1)

[0053] In the above expression (1), R is the posture, and T is the position. The posture R is a 3 x 3 matrix, and the position T is a 3 x 1 matrix. X is the three-dimensional position of the feature point Fp acquired by the detector 50, and X' is the reference position acquired by the measuring device 300. Further, for example, a mark can be given to the corners of the target TG in advance, the mark is detected, and the correspondence between the three-dimensional position of the feature point Fp and the reference position is calculated based on the positional relationship with the mark.

[0054] The calibration section 25 calibrates the posture / position of the detector 50 (step S150). Specifically, the calibration section 25 uses the detected posture / position to perform adjustment of the physical posture / position of the detector 50, or to perform calibration or alignment that corrects the detection data output from the detector 50 using the difference between the detected posture / position and the posture / position of the detector 50 decided in advance as a reference as a correction value.

[0055] The posture / position detecting apparatus 10 according to the first embodiment described above can detect the positions of the feature points Fp on the target TG with good precision because the target TG including the detection pattern KP decided in advance that is expressed on a plane is included in the object. The posture / position detecting section 23 detects the pixel positions of the feature points Fp on the target TG in the background light image HI, calculates the three-dimensional positions of the corresponding pixels Gt in the distance image KI corresponding to the pixel positions of the feature points Fp by interpolating the three-dimensional positions of the surrounding pixels Gs of the corresponding pixels Gt, that is, the surrounding point three-dimensional positions, and detects the posture / position of the detector 50 using the calculated corresponding point three-dimensional positions, and thus can calculate the three-dimensional positions of the feature points Fp with good precision compared to a configuration that detects the positions of the feature points Fp using only the distance image KI. As a result, the corresponding point three-dimensional positions can be calculated with good precision, and thus the posture / position of the detector 50 can be detected with good precision.

[0056] Second Embodiment:

[0057] The posture / position detection system of the detector 50 of the second embodiment differs from the posture / position detection system 500 of the detector 50 of the first embodiment in that the posture / position detection device 10 is mounted on the vehicle 100. The other configuration of the posture / position detection system of the detector 50 of the second embodiment is the same as that of the first embodiment, so the same components are given the same reference numerals, and detailed description thereof will be omitted.

[0058] The posture / position detection processing of the first embodiment is executed on the inspection line at the time of shipment of the vehicle 100. In contrast, the posture / position detection processing of the second embodiment is executed during running of the vehicle 100 after shipment. Therefore, in the second embodiment, the road structure set in the running environment of the vehicle 100 is used as the target TG. Hereinafter, a specific description will be given.

[0059] Figure 10 The posture / position detection processing in the second embodiment shown differs from the posture / position detection processing of the first embodiment in that step S101 and step S103 are additionally executed, and step S105a is executed instead of step S105. The other processes of the posture / position detection processing of the second embodiment are the same as those of the first embodiment, so the same processes are given the same reference numerals, and detailed description thereof will be omitted.

[0060] If the posture / position detection process is started, the posture / position detection unit 23 acquires the positions of road structures existing in the periphery of the current position of the vehicle 100 (step S101). Specifically, the posture / position detection unit 23 acquires the current position (own position) of the vehicle 100 using a position sensor not shown, which is mounted on the vehicle 100. The current position is expressed by the latitude and the longitude of the vehicle 100. As the position sensor, for example, a global navigation satellite system (GNSS), a gyro sensor, or the like can be used. Further, the current position of the vehicle 100 can be acquired via a wireless base station such as a mobile base station or a road-to-vehicle communication device provided on a road, or can be acquired using traffic information provided from VICS (Vehicle Information and Communication System / VICS is a registered trademark). Next, the posture / position detection unit 23 acquires road structures existing in the periphery of the current position of the vehicle 100 with reference to map information of a navigation device not shown, which is mounted on the vehicle 100. In the present embodiment, "road structures" mean road side objects provided on a road, a shoulder, such as curb stones, guard rails, signs, signal lights, lane dividers (lane marks), stop lines, pedestrian crossings, and the like, of a kind determined in advance. The "kind determined in advance" means road side objects that are flat in surface shape, such as signs corresponding to stop, signs corresponding to temporary stop, signal lights, lane dividers, stop lines, pedestrian crossings, and the like, which are expressed in a flat shape. Therefore, for example, signal lights with covers for snow protection, signal lights that are curved in surface shape are not included in road structures. In addition, in the present embodiment, the "map information" described above means a high-precision three-dimensional map, which includes information indicating the kind of road structures, such as the kind of signs, and the like.

[0061] The posture / position detection unit 23 sets the acquired road structures as targets TG (step S103). For example, Figure 11 The target TG1 shown is a sign of "stop" and has a detection pattern KP1. In the detection pattern KP1, Figure 11 In the drawing, symbols indicating feature points Fp5 and Fp6 are displayed on the detection pattern KP1, but the feature points Fp5 and Fp6 are not indicated on the actual detection pattern KP1, that is, the road structure. Therefore, the posture / position detection unit 23 regards the edge portion of the character of "stop" as the feature points Fp5 and Fp6 and detects the feature points Fp5 and Fp6 in the background light image HI and the distance image KI. Further, the feature points Fp are not limited to the edge portion of the character and can be set to the three-dimensional positions of the corners and the center of the sign.

[0062] The posture / position detection unit 23 acquires a reference position (step S105a). Specifically, the posture / position detection unit 23 uses map information to acquire a three-dimensional position corresponding to the feature point Fp, with the current position of the vehicle 100 as the origin, and uses this position as the reference position. That is, in the second embodiment, the reference position refers to the three-dimensional position of the feature point obtained from the high-precision three-dimensional map. After executing step S105a, step S110 described above is executed.

[0063] According to the second embodiment described above, the target TG1 is a road structure of a predetermined type that exists around the current position of the vehicle 100. The posture / position detection unit 23 uses predetermined feature points on the road structure to detect the posture / position of the detector 50, so that the posture / position of the detector 50 can be detected even when the vehicle 100 is in motion.

[0064] Third implementation method:

[0065] like Figure 12 As shown, the posture / position detection system 500a of the detector 50 in the third embodiment differs from the posture / position detection system 500 of the detector 50 in that the measuring device 300 is omitted and the vehicle 100a is provided instead of the vehicle 100. The other configurations of the posture / position detection system 500a of the detector 50 are the same as those in the first embodiment, so the same reference numerals are added to the same constituent elements, and detailed descriptions are omitted.

[0066] The vehicle 100a in the third embodiment includes a detector 60. The detector 60 is, for example, an imaging device. Alternatively, the detector 60 may replace the imaging device with a lidar, which is different from the detector 50. The detector 60 is mounted on the roof at the front of the vehicle 100a, such that at least a portion of the detection range of the detector 60 overlaps with the detection range of the detector 50. Therefore, in this embodiment, the offset of the detector 50 from a coordinate axis based on the posture / position of the detector 60 can be detected using the overlapping area (overlapping region) of the detection range of the detector 50 and the detector 60.

[0067] Figure 13 The pose / position detection process in the third embodiment shown differs from that in the first embodiment in that step S105b is performed instead of step S105. The other processes of the pose / position detection process in the third embodiment are the same as those in the first embodiment, so the same reference numerals are used for the same processes, and detailed descriptions are omitted.

[0068] If the posture detection process is started, the posture / position detection section 23 acquires the reference position (step S105b). Specifically, the posture / position detection section 23 causes the detector 60 to acquire the reference position. The detector 60 acquires the captured image of each target TG, and determines the direction and position of the reference coordinate axis. Specifically, the detector 60 extracts the pixels of the corner portion of the target TG from the captured image of each target TG using the Harris corner detection method described above, and detects the position of the feature point Fp as the reference position. That is, the reference position in the third embodiment refers to the pixel position of the feature point calculated from the captured image acquired by the detector 60. After step S105b is executed, step S110 described above is executed.

[0069] The posture / position detection apparatus 10 of the detector 50 according to the third embodiment described above has the same effects as the first embodiment described above. In addition, the posture / position of the detector 50 with respect to the detector 60 can be detected with high precision.

[0070] Other Embodiments:

[0071] (1) In each of the embodiments described above, the posture / position detection apparatus 10 can not have the calibration section 25. That is, only the posture / position of the detector 50 can be detected.

[0072] (2) In each of the embodiments described above, the detector 50 is a laser radar, but can be a TOF camera.

[0073] (3) In each of the embodiments described above, the target TG, TG1, and the detection pattern KP, KP1 are not limited to the examples described above. For example, as shown in FIG. 19, the target TG2 can have the detection pattern KP2 indicating a plurality of point patterns Pt3. In this case, the feature point Fp7 can be indicated at the center of gravity position of the point pattern Pt3. Further, for example, as shown in FIG. 20, the target TG3 can have the detection pattern KP3 indicating a plurality of AR markers Pt4. The AR marker Pt4 can be, for example, an ArUco marker. In this case, the feature point Fp8 can be indicated at the four corners of the AR marker Pt4. Figure 14 Figure 15 Further, for example, as shown in FIG. 21, the target TG4 can have the detection pattern KP4 indicating a plurality of QR codes Pt5. In this case, the feature point Fp9 can be indicated at the four corners of the QR code Pt5.

[0074] Further, for example, as shown in FIG. 22, the target TG5 can have the detection pattern KP5 indicating a plurality of QR codes Pt6. In this case, the feature point Fp10 can be indicated at the four corners of the QR code Pt6. Figure 16 ​As illustrated, the colors of the detection patterns KP can also be made different for each target TG. Specifically, the first target TG4 is provided with a detection pattern KP4, the second target TG5 is provided with a detection pattern KP5, and the third target TG6 is provided with a detection pattern KP6. Each of the detection patterns KP4, KP5, and KP6 is a checkered pattern composed of a plurality of patterns in a lattice shape. The black patterns Pt5a, Pt5b, and Pt5c in the detection patterns KP4, KP5, and KP6 are different in color from one another, and the color of the pattern Pt5a of the detection pattern KP4 is lighter than the colors of the pattern Pt5b of the detection pattern KP5 and the pattern Pt5c of the detection pattern KP6. In addition, the color of the pattern Pt5b of the detection pattern KP5 is lighter than the color of the pattern Pt5c of the detection pattern KP6. That is, the luminance values of the respective patterns Pt5a, Pt5b, and Pt5c increase in the order of the pattern Pt5a, the pattern Pt5b, and the pattern Pt5c.

[0075] Therefore, even in a case where three targets TG4, TG5, and TG6 are represented in one background light image HI, it is possible to distinguish the respective pixel values of the background light image HI by using the respective luminance values of the patterns Pt5a, Pt5b, and Pt5c as threshold values, thereby distinguishing and detecting the feature points Fp in the respective targets TG4, TG5, and TG6. In addition, for example, instead of the colors of the detection patterns KP, or in addition to the colors, the materials of the detection patterns KP can be made different. Specifically, the detection patterns KP4, KP5, and KP6 can also be provided with a reflector function, and the reflectivity of each of the detection patterns KP4, KP5, and KP6 can be made different. In addition, for example, dedicated targets TG corresponding to the types of the detectors 50 can be used. Specifically, in a case where the detector 50 is composed of a laser radar, a target TG having a reflector function for a laser radar can be used. In a case where the detector 50 is composed of a camera, a target TG provided with a detection pattern KP composed of colors capable of performing edge detection with higher precision can be used.

[0076] (4) In each of the above embodiments, the target TG is a flat panel, but can also be a cubic shape, or a cubic shape composed of flat panels combined in a concave-convex shape. Specifically, as illustrated, Figure 17 the target TG7 is a cube composed of two cubes Rs1 and Rs2 of different sizes arranged alternately. The face S1 of the target TG7 is in a concave-convex shape by alternately arranging the face S1a of the first cube Rs1 and the face S1b of the second cube Rs2. In this case, the intersection of the face S1a as a convex face and the face S1b as a concave face in the face S1 can be used as a feature point Fp. In such a structure, the position of the feature point Fp can also be detected by extracting the edges of the convex face S1a and the concave face S1b.

[0077] (5) In each of the above-described embodiments, the resolution of the background light image HI is higher than the resolution of the distance image KI, but the resolutions of the two images can be the same. By using the background light image HI to detect the feature points Fp, it is possible to set more targets to the targets. Specifically, in the case of using only the distance image KI and the reflection intensity image RI, it is necessary to detect a point of change in distance, a point of change in material of the target TG as the feature points. In contrast, in the case of using the background light image HI, it is also possible to adopt an object whose shape is flat, an object whose material is uniform, such as a sign, a billboard, a road surface, and a sign painted on a road, as a target, as long as the textures are different.

[0078] The control section and each of the other sections according to the present disclosure and the methods thereof can also be realized by a special-purpose computer provided by a processor programmed to execute one or more functions embodied by a computer program and a memory. Alternatively, the control section and each of the other sections according to the present disclosure and the methods thereof can also be realized by a special-purpose computer provided by a processor composed of one or more special-purpose hardware logic circuits. Alternatively, the control section and the methods thereof according to the present disclosure can also be realized by one or more special-purpose computers composed of a combination of a processor programmed to execute one or more functions and a processor composed of one or more hardware logic circuits. In addition, the computer program can be stored as instructions executed by a computer in a non-transitory tangible recording medium readable by the computer.

[0079] The above describes the present disclosure based on the embodiments and the modified examples, but the above-described embodiments of the invention are embodiments for making the understanding of the present disclosure easy, and do not limit the present disclosure. The present disclosure can be changed and improved without departing from the gist thereof and the scope of the claims, and includes equivalents thereof. For example, the technical features in the embodiments and the modified examples corresponding to the technical features in each of the modes recited in the summary of the invention can be appropriately replaced, combined, in order to solve part or all of the above-described problems, or in order to achieve part or all of the above-described effects. In addition, if the technical features are not described as essential technical features in the specification, they can be appropriately deleted.

Claims

1. A posture / position detection device of a detector, which is a posture / position detection device of a detector mounted on a vehicle, comprising: an image acquisition unit that acquires a distance image and a background light image from the detector, the distance image representing distances to an object within an irradiation range of irradiation light irradiated from the detector, the background light image being a background light image representing a light receiving intensity of an ambient light, and having a resolution higher than that of the distance image; and a posture / position detection unit that detects a posture / position of the detector using the distance image and the background light image, the object including a target used for the detection of the posture / position, the target having a predetermined detection pattern represented on a plane, the posture / position detection unit including: a feature point detection unit that detects a pixel position of a feature point on the target in the background light image; a corresponding point three-dimensional position acquisition unit that acquires a corresponding point three-dimensional position based on a pixel value of a corresponding pixel corresponding to the pixel position of the feature point in the distance image; and a posture / position detection unit that detects the posture / position of the detector using the corresponding point three-dimensional position.

2. The posture / position detection device of a detector according to claim 1, wherein the resolution of the background light image is higher than the resolution of the distance image.

3. The posture / position detection device of a detector according to claim 1, wherein the posture / position detection unit acquires a surrounding point three-dimensional position based on each pixel value of surrounding pixels around the corresponding pixel in the distance image, and calculates the corresponding point three-dimensional position by interpolating the surrounding point three-dimensional position.

4. The posture / position detection device of a detector according to claim 1, wherein the target includes a plurality of targets different from each other in at least one of color and material.

5. The posture / position detection device of a detector according to claim 3, wherein the target includes a plurality of targets different from each other in at least one of color and material.

6. The posture / position detection device of a detector according to any one of claims 1 to 5, wherein the target is a predetermined kind of road structure existing in a periphery of a current position of the vehicle, and the posture / position detection unit detects the posture / position of the detector using a predetermined feature point on the road structure.

7. The posture / position detection device of a detector according to any one of claims 1 to 5, further comprising a calibration unit that calibrates the detector using the detected posture / position of the detector.

8. The posture / position detection device of a detector according to any one of claims 1 to 5, wherein the detector has a plurality of light receiving elements that can receive incident light including reflected light of the irradiation light, and the light receiving element is a single photon avalanche diode that outputs a light receiving value corresponding to a light receiving state of the light receiving element.

9. The posture / position detection device of a detector according to any one of claims 1 to 5, wherein the detector is a laser radar. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ 10. A posture / position detection method of a detector, which is a posture / position detection method of a detector mounted on a vehicle, a distance image representing a distance to an object within an irradiation range of irradiation light irradiated from the detector, and a background light image representing a light receiving intensity of an environmental light, the resolution of the background light image being higher than the resolution of the distance image, the posture / position of the detector is detected using the distance image and the background light image, the object includes a target, the target being a target used for the detection of the posture / position, and having a predetermined detection pattern represented on a plane, a pixel position of a feature point on the target in the background light image is detected, a corresponding point three-dimensional position is acquired from a pixel value of a corresponding pixel, the corresponding pixel being a pixel in the distance image corresponding to the pixel position of the feature point, the posture / position of the detector is detected using the corresponding point three-dimensional position.

11. The posture / position detection method of a detector according to claim 10, wherein the resolution of the background light image is higher than the resolution of the distance image.

Images